List view
Habitual Capture. Make logira easy to use continuously with AI coding agents. This release focuses on reducing the friction of repeatedly wrapping Claude/Codex commands. Users should be able to configure local defaults and use shell integration or command shims so that common AI agent commands are captured automatically. Must-have: - shell integration or shim-based auto-capture for AI agent commands - minimal YAML config for runtime defaults and auto-capture targets Should-have: - explicit run tracking modes for background/descendant process behavior Non-goals: - fully transparent system-wide process monitoring - hosted UI - cloud sync - team/fleet management - blocking or enforcement
No due date•0/3 issues closedReadable Runs. Make a single `logira run` immediately understandable. This release focuses on improving the post-run experience for local AI coding agent audits. After running Claude/Codex under logira, users should quickly see what happened, whether any detections were triggered, and what to inspect next. Must-have: - end-of-run summary with event counts, detection counts, simple risk level, and next-command hints Should-have: - shareable local report export - improved AI agent quickstart docs Non-goals: - auto-capture / watch mode - hosted UI - cloud sync - team policy management - blocking or enforcement
No due date•1/3 issues closed